Performance assessment of Tsunami-HySEA model for NTHMP tsunami currents benchmarking. Field cases

Highlights

• Assessment of the Tsunami-HySEA model for tsunami currents is performed.

• Measured observed data from the 2011 Tohoku tsunami are used for verification.

• The field data used were proposed by NTHMP for tsunami-currents model validation.

• Hydrostatic version of the numerical model Tsunami-HySEA is used.

• Model results are in good agreement with observed field data.

Abstract

In this study, we present model results derived from the Tsunami Current Benchmarking Workshop held by the NTHMP (National Tsunami Hazard Mitigation Program) at Portland, Oregon, US, in February 2015 for the two benchmark problems dealing with field data. In this workshop, the Tsunami-HySEA model was used to perform the five proposed numerical benchmarks. The other three benchmarks, which are based on lab data, are the subject of a related paper (Macías et al., 2020). The results and comparisons with measured data for the two benchmarks which deal with field observations, described in the present paper, include: (1) Benchmark #2, impact of tsunami waves at Hilo Harbor, Hawaii, and (2) Benchmark #3, tsunami currents at Tauranga Harbor, New Zealand, both for data of the 2011 Tohoku tsunami. The numerical results demonstrate that the Tsunami-HySEA model is able to accurately reproduce sea surface elevations and current velocities for both the benchmark problems performed in a reduced local domain as in the case of the complete scenario modeled from the source to Hilo Harbor on the Pacific Ocean.

Introduction

The tsunami-modeling community has focused most of its effort on modeling wave height and runup as the key components of tsunami hazard assessment, and model validation has been mostly restricted to these variables (Nicolsky et al., 2011, NTHMP, 2012 and references therein; Tolkova, 2014, Horrillo et al., 2015, and many others). On the other hand, studies focused on modeling the effects of tsunami currents or devoted to model validation for tsunami currents are less common (Lynett et al., 2014). Nevertheless, the impact caused by the strong currents that typically accompany a tsunami is of key importance in ports, bays, and harbours. It is even important with regard to streets in between buildings for larger tsunamis (as in Onagawa during the 2011 Tohoku tsunami). Besides water height, the ability to compute and predict tsunami flow velocities has revealed of great importance in risk assessment and hazard mitigation. Model results are increasingly being used to estimate damage to coastal infrastructure, and therefore understanding the accuracy and precision of velocity predictions for tsunami currents becomes of undeniable importance. Nonetheless, until recently, a benchmarking based on velocity current observations was difficult to perform as few direct measurements of tsunami velocities existed to compare with the model results. This has changed since the 2011 Tohoku tsunami as many current-meters were deployed in several locations around the Pacific. However, despite the deployment of these current-meters, the numerical results and comparisons presented here and during the workshop, revealed some deficiencies in these observed data (in particular in the frequency of sampling) which do not allow a completely satisfactory comparison with the numerical results.

Being aware of this need for validation and verification, and also aware of the lack of model benchmarking for tsunami currents, the National Tsunami Hazard Mitigation Program (NTHMP) decided to implement a benchmarking procedure to evaluate NTHMP-approved models for current predictions. The previous benchmark tests mandated by the NTHMP focused on wave height and runup (NTHMP, 2012). To fulfill this new mandate a benchmarking workshop was organized in February, 2015 in Portland, Oregon (http://coastal.usc.edu/currents_workshop/), called “NTHMP/MMS Benchmarking Workshop: Tsunami currents”. For this workshop, five different benchmark problems were proposed, including lab-measured datasets and observed data. The first two benchmark problems were mandatory for all participating modelers/models, with the remainder of the tests being optional. Benchmarks #1, #4, and #5 were based on lab-measured data, and are the subject of another paper (Macías et al., 2020). BP2 proposed the simulation of a field case for the Hilo Harbor (Hawaii, USA) during the 2011 Tohoku tsunami. The third test dealt with the impact of the 2011 Tohoku tsunami on Tauranga Harbor, New Zealand, with the aim of including the effects of tides. The EDANYA group (www.uma.es/edanya) from the University of Málaga participated in the workshop with the Tsunami-HySEA model and presented numerical results for all five of the benchmark problems proposed. As output of the workshop, a model inter-comparison study was elaborated (Lynett et al., 2017) and a report containing results from all models was produced (NTHMP, 2016). In the present work, the numerical results obtained with the Tsunami-HySEA model for the two benchmark problems based on field data are described and compared by the provided observed data.

This paper is outlined as follows. Section 2 describes the numerical model: equations and numerical aspects. Sections 3 Benchmark problem #2: Tsunami currents in Hilo Harbor, 4 Benchmark problem #3: Tsunami currents in Tauranga Harbor describe the two benchmark problems and present the numerical results obtained with the Tsunami-HySEA model, for Hilo Harbor and Tauranga Harbor, respectively. Section 5 concludes the paper with some final comments and remarks.